Skip to main content
Glama
TMHSDigital

Screencast MCP

by TMHSDigital
OverviewSchemaSourceRelated ServersDiscussionsAdmin
TypeScript
Local

start_recording

Starts a background screen recording of the full desktop, a monitor, window, or region and returns a session ID and output path. Stop it with stop_recording to finalize the video.

Instructions

Start a screen recording (Windows gdigrab) as a background ffmpeg process. Returns a session id and output path. Recording is explicit and never auto-starts. Use stop_recording with the id to finalize the file. Audio is video-only by default; set audio.source = 'system' to also capture system (loopback) audio. Microphone capture is not supported.

Input Schema

NameRequiredDescriptionDefault
fpsNoFrames per second (default 30).
keysNoWith events: masked (default) logs typing only as bursts (start and length on a 0.5 s grid), never which keys or when each one was pressed. shortcuts also names chords made with Ctrl, Alt, or Win (e.g. Ctrl+S), with their times; typed text is never logged.
audioNoAudio capture (default none, video-only). source 'system' needs a loopback device; microphone capture is not supported.
eventsNoAlso log mouse clicks, scroll, cursor positions, and typing to <recording>.events.jsonl, timed to the video and in video pixels (default false). Read it with read_events. Adds about a second to start.
outputNoOptional output .mp4 path. Defaults to a file under SCREENCAST_HOME/recordings.
targetYesCapture target: 'full' | 'monitor:<index>' (0 = primary) | 'window:<title>' (or window:exe=<app.exe> / window:hwnd=<handle>, without relying on the title) | 'region:<x>,<y>,<w>,<h>' (absolute pixels). window: records the on-screen rectangle the window occupies, resolved ONCE at start (it must be visible, on top, not minimized); it does not follow a window moved or resized mid-recording.
backendNogdigrab (default) records the screen rectangle. wgc (window: targets, FFmpeg 8+) records the window's own surface through Windows Graphics Capture: it keeps recording the window when it is covered, moved, or in the background, and frames only its client area. No audio with wgc.
qualityNoEncoder preset (default standard).
overwriteNoAllow replacing an existing file at the output path (default false).
exactTitleNoFor window: targets, match only a window whose title equals the given title (case-insensitive); never fall back to a substring match. Default false.
maxDurationSecNoStop automatically after this many seconds, finalizing the file (default 3600, or SCREENCAST_MAX_RECORDING_SEC; at most 604800, 7 days). 0 = no cap. A safety net against a recording that is never stopped.

Schema Changelog

Changes observed during successful MCP inspections.

  1. Changed4 schema fields changedv0.39.0
    • changedInput schema / properties / events / description
      Previous value: -"Also log mouse clicks, scroll, cursor positions, and key presses to <recording>.events.jsonl, timed to the video and in video pixels (default false). Read it with read_events. Adds about a second to start."New value: +"Also log mouse clicks, scroll, cursor positions, and typing to <recording>.events.jsonl, timed to the video and in video pixels (default false). Read it with read_events. Adds about a second to start."
    • changedInput schema / properties / keys / description
      Previous value: -"With events: masked (default) logs that a key was pressed, never which. shortcuts also names chords made with Ctrl, Alt, or Win (e.g. Ctrl+S); typed text is never logged."New value: +"With events: masked (default) logs typing only as bursts (start and length on a 0.5 s grid), never which keys or when each one was pressed. shortcuts also names chords made with Ctrl, Alt, or Win (e.g. Ctrl+S), with their times; typed text is never logged."
    • changedInput schema / properties / maxDurationSec / description
      Previous value: -"Stop automatically after this many seconds, finalizing the file (default 3600, or SCREENCAST_MAX_RECORDING_SEC). 0 = no cap. A safety net against a recording that is never stopped."New value: +"Stop automatically after this many seconds, finalizing the file (default 3600, or SCREENCAST_MAX_RECORDING_SEC; at most 604800, 7 days). 0 = no cap. A safety net against a recording that is never stopped."
    • changedInput schema / properties / maxDurationSec / maximum
      Previous value: -9007199254740991New value: +604800
  2. First observedv0.32.0

TDQS

A4.2/5.0
Behavior4/5

Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?

With no annotations, the description carries the behavioral burden and does disclose key traits: it runs a background ffmpeg process, returns a session id and output path, never auto-starts, and requires stop_recording to finalize. It is less complete on failure modes, permissions, and auto-stop behavior, but the core lifecycle is transparent.

Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.

Conciseness5/5

Is the description appropriately sized, front-loaded, and free of redundancy?

The description is compact and front-loaded: it opens with the action and implementation, then covers return values, lifecycle, finalization, and audio limitations. Every sentence adds useful information with no filler.

Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.

Completeness4/5

Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?

Given the rich input schema, the description is complete enough for an agent to understand the core operation, return contract, lifecycle, and audio constraint. It relies on the schema for parameter details and does not cover failure modes or permissions, but no critical invocation detail is missing.

Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.

Parameters3/5

Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?

Schema description coverage is 100%, so the schema already documents all 11 parameters in detail. The description reiterates the audio default and microphone limitation but adds little parameter semantics beyond what the schema provides, making 3 the appropriate baseline.

Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.

Purpose5/5

Does the description clearly state what the tool does and how it differs from similar tools?

The description states a specific verb and resource: start a screen recording using Windows gdigrab as a background ffmpeg process. It also distinguishes the tool from its natural sibling by naming stop_recording as the finalization step, so an agent can tell it apart from stop_recording or screenshot tools.

Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.

Usage Guidelines4/5

Does the description explain when to use this tool, when not to, or what alternatives exist?

The description gives clear lifecycle guidance: recording is explicit, never auto-starts, and must be finalized with stop_recording using the returned id. It also states an important exclusion, that microphone capture is not supported, though it does not explicitly compare start_recording with alternatives like screenshot for static capture.

Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.

Deploy Server

Other Tools